Inter-pathologist Agreement on Structured Histopathology Reporting in Chronic Rhinosinusitis

Abstract

Structured histopathology reporting is increasingly being utilized in rhinology to characterize endotypes in chronic rhinosinusitis and guide management decisions after sinus surgery. The goal of this investigation is to evaluate inter-observer agreement in structured histopathology reporting. Two experienced head and neck pathologists independently compiled structured histopathology reports for tissue samples collected during functional endoscopic sinus surgery. Cohen's standard kappa (κ) coefficients were calculated for each histopathologic variable to assess inter-pathologist agreement. A total of 92 cases were analyzed. Substantial inter-pathologist agreement was reached on tissue eosinophil count (κ = 0.64, P < .001), the presence of eosinophil aggregates (κ = 0.62, P < .001), and the presence of fungal elements (κ = 0.74, P < .001). There was moderate agreement on the degree of inflammation (κ = 0.56, P < .001) and the presence of squamous metaplasia (κ = 0.46, P < .001). There was fair agreement on the presence of neutrophil infiltrates (κ = 0.33, P < .001), the presence of hyperplastic changes (κ = 0.40, P < .001), and the presence of fibrosis (κ = 0.24, P = .022). There was only slight agreement on the degree of subepithelial edema (κ = 0.20, P = .008). The κ coefficients for basement membrane thickening and mucosal ulceration were not statistically significant. High inter-pathologist agreement was demonstrated for several salient histopathologic variables, including tissue eosinophil count and the presence of eosinophil aggregates. However, refining the definitions of certain histopathologic variables may improve the reproducibility of structured histopathology reporting.